Journal of Animal Ecology 2008, 77, 275–284
doi: 10.1111/j.1365-2656.2007.01336.x
Biotic disturbance and benthic community dynamics
in salmon-bearing streams
Blackwell Publishing Ltd
Jonathan W. Moore* and Daniel E. Schindler
School of Aquatic and Fishery Sciences, Box 350220, University of Washington, Seattle, WA 98195, USA
Summary
1. Organisms can impact ecosystems via multiple pathways, often with positive and negative
impacts on inhabitants. Understanding the context dependency of these types of impacts remains
challenging. For example, organisms may perform different functions at different densities.
2. Anadromous salmon accumulate > 99% of their lifetime growth in marine ecosystems, and then
return to spawn, often at high densities, in relatively confined freshwaters. While previous research
has focused on how salmon nutrients can fertilize benthic communities, we examined how an ecosystem engineer, sockeye salmon Oncorhynchus nerka, influences seasonal dynamics of stream
benthic communities through their nest-digging activities in south-western Alaska, USA. Benthic
invertebrate and algal abundance were quantified every 7–14 days during the open water seasons
of 10 streams in riffle and run habitats across multiple years, leading to 25 different stream-year
combinations that spanned a large gradient of salmon density.
3. In streams with few or no salmon, benthic algal and insect biomass were fairly constant throughout the season. However, in streams with more than 0·1 salmon m–2, algal and insect biomass
decreased by an average of 75–85% during salmon spawning. Algal biomass recovered quickly following salmon disturbance, occasionally reaching pre-salmon biomass. In contrast, in streams with
more than 0·1 salmon m–2, aquatic insect populations did not recover to pre-salmon levels within
the same season. We observed no positive impacts of salmon on algae or insects via fertilization
from carcass nutrients.
4. Salmon, when their populations exceed thresholds in spawning density, are an important component of stream disturbance regimes and influence seasonal dynamics of benthic communities.
Human activities that drive salmon densities below threshold densities, as has likely happened in
many streams, will lead to altered seasonal dynamics of stream communities. Human activities that
alter animal populations that are sources of biogenic disturbance can result in shifts in community
dynamics.
Key-words: bioturbation, ecosystem engineer, marine-derived nutrients, Oncorhynchus nerka,
succession.
Introduction
Organisms can impact ecosystems via multiple pathways,
often with positive and negative impacts on coinhabitants.
For example, many plants act as competitors under some
circumstances or scales, even while acting as facilitators under
other conditions (Callaway & Walker 1997). Alternatively,
Prochilodus, a dominant detritivorous fish, has been found to
alter tropical stream ecosystems via several pathways. Largeand small-scale experiments have demonstrated that through
foraging on detritus and perturbing benthic habitats, Prochilodus decrease sediment accrual rates and periphyton abun*Correspondence author. E-mail: Jonathan.Moore@noaa.gov
dance (Flecker 1997), decreasing primary productivity and
altering carbon cycling (Taylor, Flecker & Hall 2006). However,
McIntyre et al. (2007) predicted that these fish could positively
impact primary producers by excreting large amounts of
limiting nutrients. Reconciling the different pathways by which
organisms alter their ecosystems under different contexts is
particularly challenging given the increasing realization that
organisms can exert large impacts via both trophic (e.g. predation)
and nontrophic interactions (e.g. ecosystem engineering)
(Bruno, Stachowicz & Bertness 2003; Wright & Jones 2006).
One aspect of context dependency is the functional relationship between the density of a species and its impact on its environment. For example, the relationships between densities of
bioturbators, a type of ecosystem engineer, and their impacts
© 2007 The Authors. Journal compilation © 2007 British Ecological Society
276
J. W. Moore & D. E. Schindler
on aquatic ecosystems are relatively understudied (Moore
2006). Understanding these relationships is increasingly
important in light of concerns regarding the ecological consequences of declines in biomass of benthic bioturbators due
to anthropogenic impacts (e.g. Coleman & Williams 2002;
Taylor et al. 2006). While a recent simulation analysis assumed
that impacts of bioturbators increase linearly with population
density (Solan et al. 2004), it seems likely that impacts of these
organisms may often be a nonlinear function of population
densities. For example, Flecker, Feifarek & Taylor (1999)
found a saturating relationship between tadpole density and
sediment accumulation – at five tadpoles per square metre,
sediment accumulation ceased decreasing as a function of
increasing tadpole density. Identifying these relationships will
aid understanding when organisms contribute substantially
to disturbance regimes, and subsequently understanding how
human activities may change disturbance regimes.
Pacific salmon Oncorhynchus spp. may exert strong impacts
on coastal stream communities via several pathways. Anadromous and semelparous salmon accumulate over 99% of their
mass during the ocean phase of their life history, and then
return to coastal freshwaters to spawn and die (Quinn 2005).
Through foraging in the productive and vast oceanic ecosystem, salmon can reach large sizes and spawn at high densities
relative to the confined coastal freshwaters in which they spawn.
Salmon dig large nests during spawning, perturbing benthic
habitats of spawning reaches (Montgomery et al. 1996;
Gottesfeld et al. 2004). Indeed, an experimental exclusion of
salmon suggested that salmon nest digging decreases fine
sediment accumulation and algal biomass (Moore, Schindler
& Scheuerell 2004). In addition, several studies have observed
decreases in algal and benthic insect biomass during salmon
spawning in a single or a few local areas (e.g. Hildebrand 1971;
Field-Dodgson 1987; Peterson & Foote 2000; Minakawa &
Gara 2003; Lessard & Merritt 2006). While these studies have
documented substantial local impacts of salmon nest digging
on streams, the relationship between salmon population density
and their impact on benthic communities remains unknown.
In addition, previous research on the impacts of salmon on
streams has emphasized how nutrients from salmon carcasses,
derived from oceanic ecosystems due to their anadromous life
history, may subsidize streams from the bottom-up (sensu
Polis, Anderson & Holt 1997; reviewed by Gende et al. 2002;
Naiman et al. 2002). Specifically, experimental additions of
salmon carcasses have suggested that nutrients from salmon
can fertilize stream food webs, increasing algal and benthic
invertebrate biomass (e.g. Wipfli, Hudson & Caouette 1998;
Chaloner & Wipfli 2002) and comparative studies have observed
increased nutrients and algae in response to spawning salmon
(e.g. Johnston et al. 2004). Thus, it is unknown whether
salmon have net positive (fertilization via salmon-derived
nutrients) or net negative (disturbance from nest-digging)
impacts on stream benthic communities, and how or if these
roles change as a function of salmon density.
The objective of this study was to examine how different
densities of salmon alter the seasonal dynamics of stream
benthic communities. By quantifying the seasonal dynamics
Fig. 1. Map of study locations, in the Wood River system of the
Bristol Bay region of south-eastern Alaska. Streams are highlighted
by bold. Inset demonstrates the location of this system in the state of
Alaska, USA.
of benthic algae and invertebrates in at least 25 stream-year
combinations, we show for the first time that above a threshold
density of salmon, bioturbation by nest-digging salmon consistently causes drastic decreases in algal and insect biomass.
We did not observe any fertilization impacts of salmon on
benthic communities at any densities. This study demonstrates
that salmon are an important component of stream disturbance regimes, and that population declines of salmon likely
change the fundamental disturbance regimes characteristic
of coastal stream ecosystems.
Methods
STUDY SITE
This study was based on streams in the Wood River system of southwestern Alaska, USA. This 300 000 ha river system is a spawning
and nursery habitat for anadromous sockeye salmon (O. nerka). An
average of 1·1 million sockeye spawn in the creeks, rivers and
beaches of this system every year. An additional 1·8 million sockeye
on average are harvested every year prior to spawning (Baker et al.
2006). We quantified how salmon altered seasonal dynamics of
stream benthic communities for up to five summers on 10 streams
for a total of 25 and 27 stream-year combinations for benthic invertebrate and periphyton, respectively (Table 1; Fig. 1). Streams
spanned a natural gradient in salmon densities (0–0·54 salmon m–2).
This gradient in salmon densities was achieved through year to year
variation in salmon returns and differences in densities among
streams (Table 1). A previous survey of our study streams found
that much of their total habitat was suitable for salmon spawning
(Marriott 1964). Based on this survey, spawning habitat comprised
the following proportions of total stream areas: Bear = 0·85,
Elva = 0·15, Hansen = 1·0, Hidden = 0·87, Lynx = 0·6, Pick = 0·9,
© 2007 The Authors. Journal compilation © 2007 British Ecological Society, Journal of Animal Ecology, 77, 275–284
Biotic disturbance and stream community dynamics 277
Table 1. Streams and years sampled for the seasonal dynamics of
benthic communities. Unless otherwise noted, both benthic invertebrates and algal dynamics were quantified for these stream-years.
Salmon density represents the maximum observed live salmon
density. Flow data represent the maximum and minimum observed
flows and the dates of those observations for each stream-year and
are from Moore et al. (2007)
Flow [m3 s–1 (date)]
Stream
Year
Salmon m–2
Minimum
Maximum
Bear
2004*
2005†
2002
2003
2004
2005
2002
2005
2004*
2005†
2001
2002
2003
2004
2005
2002
2003
2004
2005
2003
2004
2005
2001
2002
2003
2004
2005
2004*
2005†
2004*
2005†
0·20
0·25
0·00
0·00
0·00
0·01
0·05
0·15
0·04
0·02
0·12
0·13
0·27
0·20
0·19
0·06
0·07
0·54
0·25
0·00
0·16
0·00
0·16
0·14
0·16
0·33
0·30
0·01
0·06
0·36
0·25
0·12 (8/20)
0·09 (8/02)
0·02 (9/03)
0·10 (8/05)
0·0 (8/04)
0·04 (8/16)
0·64 (9/03)
1·02 (8/16)
0·05 (7/25)
0·03 (8/09)
NA
0·13 (9/03)
0·16 (8/05)
0·29 (8/18)
0·14 (8/16)
0·31 (8/06)
0·37 (7/21)
0·31 (9/08)
0·17 (8/14)
0·05 (8/06)
0·06 (7/24)
0·04 (7/19)
NA
0·26 (9/06)
0·28 (7/27)
0·24 (7/24)
0·44 (8/12)
0·18 (8/09)
0·16 (8/15)
0·20 (7/20)
0·10 (8/09)
0·60 (6/17)
1·01 (8/24)
1·18 (6/23)
3·17 (6/26)
1·86 (6/27)
0·74 (6/21)
3·15 (6/23)
8·81 (6/21)
0·13 (6/16)
0·09 (6/15)
NA
0·64 (6/25)
0·77 (6/21)
0·59 (6/25)
0·71 (6/21)
1·03 (6/22)
2·48 (8/23)
1·20 (6/12)
1·41 (6/16)
0·09 (6/26)
0·10 (6/12)
0·08 (9/23)
NA
1·14 (6/24)
1·41 (8/21)
1·18 (6/11)
1·49 (6/16)
0·57 (6/16)
0·37 (6/25)
0·48 (6/16)
0·61 (6/15)
Cottonwood
Elva
Hansen
Hidden
Lynx
N-4
Pick
Whitefish
Yako
NA, cases where data are not available.
*Only benthic invertebrates, not algae, were quantified this
stream-year.
†Only algae, not benthic invertebrates, were quantified this
stream-year.
Whitefish = 0·75, Yako = 0·5; note that Cottonwood and N-4 were
not surveyed in this study. Across all of our study streams, salmon
were almost exclusively sockeye (> 98% of total counted salmon).
Water flows generally were highest in the early spring associated
with snowmelt or in the late summer associated with fall rainstorms
(Table 1). Flows tended to be lowest during July and early August,
the period of peak salmon spawning. We monitored benthic communities every 11·3 ± 4·3 days (mean ± 1 SD) from June up until late
August or September in 2001–05. This seasonal period spanned the
bulk of the ice-free period, and included times before, during and
after salmon spawning. At each sampling period we estimated
salmon densities in each stream by visually enumerating salmon and
dividing these counts by the wetted channel area. These salmon density estimates were performed on the stretch of stream between the
mouth of the stream and the uppermost sampling location located
near the upstream extent of salmon spawning in each stream. Maximum observed live salmon densities for each stream-year were used
as an index of salmon abundance.
PERIPHYTON
During each sampling period we quantified periphyton biomass
(chlorophyll-a cm–2) on naturally occurring rocks in our study
streams. On average we quantified algae on 5·3 ± 0·9 rocks (± 1 SD)
per stream per sampling date. Each rock was collected haphazardly
from a different sampling location. Sampling locations were selected
to maximize coverage of the length of the stream accessible to salmon,
were located in runs, and were held constant throughout the length
of the study. We sampled rocks that were small cobble (maximum
surface area was 252 cm2). Rocks were scrubbed with a small brush,
and vigorously rinsed to remove all benthic algae. We filtered a known
fraction of the filtrate through 47 mm GF/F silica filters (pore
size = 0·7 µm). The filters were frozen at –10 °C and later solubilized
in methanol. Subsequently, we determined the concentrations of
chlorophyll-a with fluorimetry (Steinman, Lamberti & Leavitt
2006). To calculate the amount of chlorophyll-a on an areal basis
(µg cm–2), we estimated the surface area of each collected rock by
assuming it was a rectangular box and measuring its average length,
width and height. The average rock was 102·8 ± 38·8 cm2 (± 1 SD;
n = 1186).
BENTHIC INVERTEBRATES
Benthic invertebrate communities were also quantified during the
open water seasons of streams, concurrent with periphyton sampling.
At each sampling event we collected 4·8 ± 1·3 Surber samples (± 1
SD; 0·5-mm mesh; sampling to a depth of 10 cm; n = 1063), each from
a different sampling location in the stream (see above for description of sampling locations). Samples were preserved in 70% ethanol.
Invertebrates were sorted and identified to the taxonomic resolution
listed below (Merritt & Cummins 1996; Stewart & Stark 2002; Adams
2004). For each taxa in each sample, the body length of five to 10
random individuals was measured using an ocular micrometer. Body
lengths were converted to dry masses using taxon specific length–
mass relationships determined for these streams (J.W. Moore,
unpublished data). Total dry mass of each taxonomic group in each
sample was estimated by multiplying abundance by the average
mass of an individual in that sample. For the purposes of this study,
we combined invertebrate taxa into the following five categories
corresponding to order or higher: Ephemeroptera (primarily Heptageneidae, Baetidae and Ephemerellidae), Plecopterans (primarily
Chloroperlidae, Nemouridae and Perlolidae), Tricopterans (primarily
Brachycentridae and Limnephilidae), Dipterans (primarily Chironomidae, Tipulidae, Simuliidae and Cerpogonidae) and other (primarily
Turbellaria, Acarina and Oligochaeta).
STATISTICAL ANALYSES
All statistical analyses were performed using  11·0. For most
analyses on algae and invertebrate dynamics, we used each streamyear as the unit of replication. The possibility that average spring
algal biomass is influenced by salmon abundance the previous year
was statistically examined using a General Linear Model (GLM)
with salmon density in the previous year and stream as potential
factors. For this analysis, average spring algal biomass was cube
root transformed to improve normality. On a finer time-scale, a
© 2007 The Authors. Journal compilation © 2007 British Ecological Society, Journal of Animal Ecology, 77, 275–284
278 J. W. Moore & D. E. Schindler
Fig. 2. Seasonal dynamics of algal biomass
in a representative stream with, and without,
high densities of salmon, across years. Filled
circles with error bars represent the mean ± 1
standard error of the mean (SEM) of algal
biomass, as quantified by chlorophyll-a.
Salmon densities are represented by shaded
areas that correspond to the right y-axis.
Note the different scales on the left y-axis.
regression was run to investigate how salmon density impacted
average algal biomass for a given date. For this analysis, algal data
was ln-transformed to improve normality. In addition, for each
stream-year, we estimated the proportional change in algal and
invertebrate biomass from the spring to salmon spawning period.
Proportional change was calculated as the proportional difference
between average spring algal biomass and algal biomass during
maximum salmon density. For streams without salmon, the period
corresponding to maximum salmon densities occurred at the beginning of August. We fit a power function to the relationship between
spawner density and the seasonal change in algal biomass. The same
function was used to fit relationships between spawner densities and
seasonal changes in biomass of invertebrate taxa. In addition, to
explicitly compare across taxa, we investigated how invertebrate
biomass changed through the season by comparing proportional
change in stream-years with low salmon densities (< 0·1 salmon m–2)
to stream-years with high salmon densities. This threshold was
chosen based on data inspection. We then ran a two-factor 
with taxa and salmon density (high vs. low) as potential effects. This
 was run once for each taxa, considering each stream-year
to be a replicate. Caddisflies were excluded from these analyses
because they were often at low densities, rendering poor estimates
of changes in proportional change. Following a significant result,
we ran post-hoc paired t-tests to examine how different taxa
responded differently to salmon with an adjusted α for multiple
comparisons.
Results
IMPACTS OF SALMON ON PERIPHYTON
Salmon nest-digging strongly impacted algal dynamics in
streams with high densities of salmon. Specifically, algal biomass tended to increase through the spring, but then decrease
as salmon entered streams and initiated spawning (Fig. 2).
For example, in Pick Creek in 2004, algal biomass increased
through the spring, but then decreased 94% when salmon first
entered the stream. Throughout the salmon spawning period
that lasted approximately a month, algal biomass was maintained at low levels. When all salmon in Pick Creek died, algal
biomass rebounded to pre-salmon levels. In contrast, in streams
with low densities of salmon, algal biomass tended to increase
through the year, although there was some stochasticity, perhaps
due to flow regime (see Fig. 2). For example, in Cottonwood,
the temporary drop in algal biomass was associated with a
period of extremely low flows (Table 1). In years and streams
with low or no spawning salmon (< 0·1 salmon m–2), algal biomass tended to increase 20% (± 33·3; ± 1 SEM), although there
was substantial variability in the seasonal progression in algal
biomass (Fig. 3). Above this threshold density, algal biomass
decreased 74% on average (± 5·5; ± 1 SEM) (Fig. 3). In fact,
© 2007 The Authors. Journal compilation © 2007 British Ecological Society, Journal of Animal Ecology, 77, 275–284
Biotic disturbance and stream community dynamics 279
Fig. 3. Summary of the seasonal changes in algal biomass as a
function of salmon density. Each point represents the proportional
change in algal biomass of a given stream-year, as a function of the
observed maximum live salmon density of that stream-year.
Proportional change was calculated as the proportional difference
between average spring algal biomass and algal biomass during
maximum salmon density. The dotted line through 0 represents
stream-years with no change, below this line indicates a decrease, and
above this line represents a stream-year with increased algal biomass.
For streams without salmon, the period corresponding to maximum
salmon densities occurred at the beginning of August. The line shown
represents the best fit of a power function (r2 = 0·41). Salmon density
represents the maximum observed live salmon density in the study
stream during that year. Note that the y-axis cut-off several data
points that were subsequently located at y > 1·0.
algal biomass decreased in every stream and year with salmon
densities above 0·06 salmon m–2.
There was no evidence that salmon increased algal biomass
via fertilization by carcass nutrients. While individual streams
had significantly different average spring chlorophyll-a biomass
(GLM, stream effect: F9 = 5·93; P = 0·002), the spawning density
of salmon the previous year in that stream had no significant
impact on average spring algal biomass (GLM, salmon density
effect: F1 = 1·572, P = 0·22). On a finer time-scale, salmon
density at the time of sampling was in fact significantly
negatively related to algal biomass of that date across all
sampling events and streams (regression between salmon
and algal biomass: r2 = 0·08, P < 0·001).
IMPACTS OF SALMON ON BENTHIC INVERTEBRATES
Salmon nest-digging also strongly influenced the seasonal
dynamics of benthic invertebrates. In streams with high
densities of salmon, invertebrate biomass tended to decrease
dramatically during salmon spawning. For example, in Bear
Creek 2004, mayfly biomass decreased from 59·3 to 1·86 mg
m–3, stonefly biomass decreased from 14·5 to 0·18 mg m–3, and
dipteran biomass decreased from 6·71 to 0·57 mg m–3 in the
initial 9 days of salmon spawning (Fig. 4). In streams with the
highest densities of salmon such as Yako, Bear and Pick
Creeks, benthic insects such as mayflies and stoneflies did not
recover following the conclusion of salmon spawning and
were rarely detected for the remainder of the sampling season.
In some streams and years with abundant salmon, other
invertebrates (predominately planaria) increased following
the end of salmon spawning, but invertebrate biomass postsalmon never recovered to pre-salmon levels. In contrast, in
streams with few or no salmon, invertebrate biomass stayed
fairly constant throughout the summer, although there was
expected variability (Fig. 4). For example, in all 4 years in
Cottonwood Creek salmon densities never exceeded 0·01
salmon m–2, and total invertebrate biomass, composed predominantly of mayflies, stayed relatively constant throughout
the entire summer season. No stream with low densities of
salmon exhibited the dramatic decreases in late July or early
August that were observed in streams with high densities of
salmon (Fig. 5). In other streams with intermediate or variable
sized runs, such as Lynx and Elva, invertebrate densities did
not consistently decrease when salmon entered stream or
recovered following salmon spawning.
Across all stream-years, invertebrate biomass decreased
more during late July or early August in stream-years with
high densities of salmon (> 0·1 salmon m–2) than stream-years
with low densities of salmon (, salmon effect; F1 = 33·05,
P < 0·001) (Fig. 6). However, individual taxa exhibited different
seasonal patterns (, taxa effect, F3 = 2·64, P = 0·05). In
stream-years with > 0·1 salmon m–2, caddis, mayfly, stonefly
and dipteran biomass decreased between 75 and 86% during
salmon spawning, on average, while biomass of other invertebrates, dominated by planaria, decreased 56% during salmon
spawning, which was significantly less than the four previously
mentioned taxa (P < 0·001 for all post-hoc paired t-tests).
Discussion
Salmon nest-digging intensely disturbs stream benthic communities (Figs 2, 4 and 5). This biotic disturbance often
reduced algal and benthic invertebrate biomass by 75–85%
(Figs 3 and 6). Salmon nest-digging, and the periodic postspawn digging that maintains nests free of sediments, buries
or dislodges algae and invertebrates for the duration of spawning. This bioturbation lasted over a month in some streams.
Thus, salmon dramatically reduce algal and invertebrate biomass, confirming previous experimental (Moore et al. 2004)
and observational studies (Hildebrand 1971; Field-Dodgson
1987; Minakawa & Gara 1999, 2003; Peterson & Foote 2000).
Our study also identifies, for the first time, spawner densities
necessary for salmon nest-digging to be a substantial source
of disturbance. Above 0·1 salmon m–2, salmon consistently
decreased biomass of benthic communities, but below this
threshold, had little noticeable impacts on benthic dynamics
(Figs 3 and 6). These thresholds are of special concern in conservation biology (Huggett 2005) – human impacts that push
the system or population over a threshold can lead to rapid
ecosystem shifts.
© 2007 The Authors. Journal compilation © 2007 British Ecological Society, Journal of Animal Ecology, 77, 275–284
280
J. W. Moore & D. E. Schindler
Fig. 4. Seasonal dynamics of invertebrate biomass in streams with no salmon or low salmon densities. Stacked areas correspond to biomass of
different invertebrate taxa. Live salmon densities are portrayed by open circles and thick grey lines that correspond to the right y-axis. Note the
different scales on the left y-axis.
Following conclusion of salmon spawning, algal biomass
increased, sometimes approaching pre-spawning biomass by
the end of the sampling season (Fig. 2). This recovery might
simply be natural succession, as algal recovery is often rapid
in streams following floods (e.g. Fisher et al. 1982). However,
it is possible that salmon indirectly facilitate algal succession
via two mechanisms. First, nutrients from salmon carcasses
may speed the recovery of algal communities. Previous studies
have noted that nutrients from salmon carcasses increase algal
accrual rates (e.g. Johnston et al. 2004). Indeed, a concurrent
study found that spawning salmon were strongly associated
with elevated levels of bioavailable nutrients in these streams
© 2007 The Authors. Journal compilation © 2007 British Ecological Society, Journal of Animal Ecology, 77, 275–284
Biotic disturbance and stream community dynamics 281
Fig. 5. Seasonal dynamics of invertebrate biomass in streams with high densities of salmon. Stacked areas correspond to biomass of different
invertebrate groups. Live salmon densities are portrayed by open circles and thick grey lines that correspond to the right y-axis. Note the different
scales on the left y-axis. See Fig. 4 for description of the shading patterns.
© 2007 The Authors. Journal compilation © 2007 British Ecological Society, Journal of Animal Ecology, 77, 275–284
282
J. W. Moore & D. E. Schindler
Fig. 6. Summary of the seasonal changes in invertebrate biomass as
a function of salmon density, for the five invertebrate groups. Each
point represents the proportional change in invertebrate biomass of a
given stream-year, as a function of the maximum live salmon density
of that stream-year. Proportional change was calculated as the
proportion difference between average spring invertebrate biomass
and invertebrate biomass during maximum salmon density. The
dotted line through 0 represents stream-years with no change, below
this line indicates a decrease, and above this line represents a streamyear with increased invertebrate biomass. The lines shown represent
the best fit of a power functions. For streams without salmon, the
period corresponding to maximum salmon densities was designated
as occurring at the beginning of August. Note that the y-axis cut-off
several data points that were subsequently located at y > 2·0.
(Moore et al. 2007). Second, salmon spawning might facilitate
algal recovery by decimating populations of benthic invertebrate herbivores. Benthic invertebrate biomass failed to fully
recover following disturbance from spawning salmon (Fig. 5),
and benthic invertebrate herbivores can exert strong top-down
control of periphyton in streams (e.g. Power 1990). Despite
these possibilities, we did not observe any evidence that salmon
positively impact algal biomass on long or short time frames.
Algal biomass was not positively impacted by the spawning
density the previous year. At a finer temporal resolution, increased salmon density only decreased algal biomass.
Salmon nest-digging also decreased biomass of benthic
invertebrates, controlling their seasonal dynamics. In contrast
with algal communities, invertebrate populations consistently
failed to recover from this biotic disturbance during the same
growing season. Indeed, invertebrate biomass in stream-years
with high salmon densities never recovered to pre-salmon
biomass the same year (Fig. 5). Salmon nest-digging likely
directly reduces benthic invertebrate biomass by crushing
or dislodging them (Peterson & Foote 2000). Decreases in
aquatic insect densities during salmon spawning have been
observed in both south-western (e.g. Peterson & Foote 2000)
and south-eastern Alaskan streams (e.g. Lessard & Merrit
2006). Temporary decreases in caddisfly larval densities were
associated with salmon spawning in Washington streams
(Walter, Bilby & Fransen 2006). Thus, salmon nest-digging
will modify the seasonal dynamics of benthic communities, with
potential indirect impacts on predators of aquatic invertebrates such as stream-dwelling fishes.
Our study adds to the growing perception that salmon nestdigging is an important source of disturbance for benthic
invertebrates, but that aquatic invertebrate life history mediates the impact of the disturbance. For example, it is possible
that consistent and predictable salmon nest-digging has driven
the evolution of benthic insect life-history phenologies. Predictable and severe floods can drive evolution of stream insect
phenologies (Lytle 2002). A previous small-scale experimental exclusion of spawning salmon in two of the creeks in this
study (Hidden Creek and Pick Creek) found that mayflies and
stoneflies decreased during salmon spawning, even in plots
where salmon were excluded from spawning (Moore et al.
2004). In contrast, planarians decreased in areas with spawning salmon but did not decrease in salmon exclusions plots,
likely the result of direct disturbance. These data suggested
that salmon nest-digging has direct negative local impacts on
some taxa, namely planarians, but that mayfly and stonefly
taxa disappear from streams, perhaps due to emergence into
terrestrial adults, during or before salmon spawning. These
results, coupled with the current studies’ findings that aquatic
insects only decrease in streams with many salmon, and that
only planaria recover follow salmon spawning, suggests that
some stream invertebrates may have life histories that are locally
adapted to the predictable disturbance of salmon spawning.
Previous studies have suggested this possibility that life history
may mediate susceptibility to salmon disturbance. Minakawa
& Gara (2003) observed that benthic invertebrate densities
decreased by about 85% during chum salmon spawning in
streams in Washington, but that populations of short-lived
taxa quickly recovered following this disturbance. Similarly,
Chaloner et al. (2004) and Lessard & Merritt (2006) observed
that rapidly developing benthic invertebrates, namely chironomids, tended to be more dominant in stream reaches with high
densities of salmon, as they recovered rapidly from the disturbance of their nest-digging, while longer-lived species, namely
mayflies, were only negatively impacted by salmon. Concurrent
analyses are explicitly investigating the impacts of salmon
nest-digging on stream insect life-history characteristics.
This study suggests that the impacts of salmon on streams
are likely context-dependent. In our location, we identified that
beyond threshold densities, salmon primary impact stream
benthic communities by acting as agents of disturbance.
Indeed, disturbance from salmon nest-digging reduces the
biomass of algae and invertebrates for much of the open water
season to such an extent that there is little algae and insects
remaining to capitalize on increased nutrients or carcass
resources. These results contrast with the findings of previous
studies that have primary focused on how salmon may fertilize stream food webs from the bottom up (reviewed by Gende
et al. 2002; Naiman et al. 2002). For example, Johnston et al.
(2004) observed increases in periphyton biomass on cobble in
© 2007 The Authors. Journal compilation © 2007 British Ecological Society, Journal of Animal Ecology, 77, 275–284
Biotic disturbance and stream community dynamics 283
response to spawning salmon in natural streams in southeastern Alaska. These contrasting findings highlight that these
dominant species can have multiple impacts on coastal ecosystem that are likely context-dependent, simultaneously fertilizing and disturbing stream habitats. In addition to salmon
density, factors that may influence the context dependency
include stream characteristics such as light availability, flow
regime, substrate size, background nutrient levels, as well as
the timing and species identity of the salmon run (Tiegs et al.
in press). Indeed, it has proved challenging to understand the
context dependency of the impacts of these dominant species
(e.g. Ambrose, Wilzbach & Cummins 2004).
There have been suggestions that salmon-derived nutrients
may increase productivity of juvenile salmon by fertilizing
stream food webs, leading to a positive feedback mechanism
between adult returns and juvenile production (Stockner,
Rydin & Hyenstrand 2000; Bilby et al. 2001). This perception
that salmon populations increase biomass of stream benthic
food webs is often based on small-scale carcass addition experiments (e.g. Chaloner & Wipfli 2002), which do not
incorporate disturbance from nest-digging salmon. Salmon
carcass addition and fertilization programmes, used as a
restoration tool for streams with impoverished salmon populations (Compton et al. 2006), fail to account for biogenic
disturbance from spawning salmon, and capture little reality
of how salmon impact streams via manifold pathways.
Human activities are altering disturbance regimes via multiple factors. Most of the large river systems in the world are
now dammed (Nilsson et al. 2005), altering flood regimes
(Poff et al. 1997). Floods can increase fish production, move
sediments and nutrients, and control seasonal dynamics (Bayley
1995; Poff et al. 1997). Thus, alteration of disturbance regimes
can have wide-reaching consequences to stream ecosystems
and the services they provide. Here we show that nest-digging
salmon are also an important component of coastal stream
disturbance regimes. Human activities that drive salmon
populations below threshold densities, as has likely happened
in many coastal streams (NRC 1996), will lead to fundamental
changes in the disturbance regimes and the rearrangement of
seasonal dynamics of benthic stream communities.
Acknowledgements
This is a contribution from the Alaska Salmon Program of the University of
Washington, supported by the Gordon and Betty Moore foundation, the
National Science Foundation and the Alaska salmon processors. Numerous
people assisted with field work, including Jackie Carter, Justin Fox, Amanda
Barg, Curtis Brock, Wendy Palen, Laura Payne, Mark Scheuerell, Sue Johnson,
Gordon Holtgrieve, Matt Baker, Tessa Francis and Casey Ruff. Sorting of
benthic invertebrate samples was greatly assisted by Karen Knirsken, Jared
Wills, Greg Osborn, Justin Fox, Casey Ruff, Peter Lisi and Adam Goodwin.
Earlier drafts of this manuscript were improved by Sue Johnson and two anonymous reviewers.
References
Adams, J. (2004) Guide to Pacific Northwest Macroinvertebrate Monitoring.
The Xerces Society, Portland, OR. Available online or on CD at http://
www.xerces.org/
Ambrose, H.E., Wilzbach, M.A. & Cummins, K.W. (2004) Periphyton
response to increased light and salmon carcass introduction in northern
California streams. Journal of the North American Benthological Society, 23,
701–712.
Baker, T.T., Fair, L.F., Clark, R.A. & Hasbrouck, J.J. (2006) Review of Salmon
Escapement Goals in Bristol Bay, 2006. Fisheries Manuscript no. 06–05,
Anchorage. Alaska Department of Fish and Game, Alaska.
Bayley, P.B. (1995) Understanding large river floodplain ecosystems. BioScience,
45, 153–158.
Bilby, R.E., Fransen, B.R., Walter, J.K., Cederholm, C.J. & Scarlett, W.J.
(2001) Preliminary evaluation of the use of nitrogen stable isotope ratios to
establish escapement levels for Pacific Salmon. Fisheries (Bethesda), 26, 6–
14.
Bruno, J.F., Stachowicz, J.J. & Bertness, M.D. (2003) Inclusion of facilitation
into ecological theory. Trends in Ecology and Evolution, 18, 119–125.
Callaway, R.M. & Walker, L.R. (1997) Competition and facilitation: a synthetic
approach to interactions in plant communities. Ecology, 78, 1958–1965.
Chaloner, D.T. & Wipfli, M.S. (2002) Influence of decomposing Pacific salmon
carcasses on macroinvertebrate growth and standing stock in southeastern
Alaska streams. Journal of the North American Benthological Society, 21,
430–442.
Chaloner, D.T., Lamberti, G.A., Merritt, R.W., Mitchell, N.L., Ostrom, P.H. &
Wipfli, M.S. (2004) Variation in responses to spawning Pacific salmon
among three south-eastern Alaska streams. Freshwater Biology, 49, 587–599.
Coleman, F.C. & Williams, S.L. (2002) Overexploiting marine ecosystem
engineers: potential consequences for biodiversity. Trends in Ecology and
Evolution, 17, 40–44.
Compton, J.E., Andersen, C.P., Phillips, D.L., Brooks, J.R., Johnson, M.G.,
Church, M.R., Hogsett, W.E., Cairns, M.A., Rygiewicz, P.T., McComb,
B.C. & Shaff, C.D. (2006) Ecological and water quality consequences of
nutrient addition for salmon restoration in the Pacific Northwest. Frontiers
in Ecology and the Environment, 4, 18–26.
Field-Dodgson, M.S. (1987) The effect of salmon redd excavation on stream
substrate and benthic community of two salmon spawning streams in Canterbury, New Zealand. Hydrobiologia, 154, 3–11.
Fisher, S.G., Gray, L.J., Grimm, N.B. & Busch, D.E. (1982) Temporal succession
in a desert stream ecosystem following flash flooding. Ecological Monographs,
52, 93–110.
Flecker, A.S. (1997) Habitat modification by tropical fishes: Environmental
heterogeneity and the variability of interaction strength. Journal of the North
American Benthological Society, 16, 286–295.
Flecker, A.S., Feifarek, B.P. & Taylor, B.W. (1999) Ecosystem engineering by a
tropical tadpole: density-dependent effects on habitat structure and larval
growth rates. Copeia, 1999, 495–500.
Gende, S.M., Edwards, R.T., Willson, M.F. & Wipfli, M.S. (2002) Pacific
salmon in aquatic and terrestrial ecosystems. Bioscience, 52, 917–928.
Gottesfeld, A.S., Hassan, M.A., Tunnifcliffe, J.F. & Poirier, R.W. (2004)
Sediment dispersion in salmon spawning streams: the influence of floods
and salmon redd construction. Journal of the American Water Resources
Association, 40, 1071–1086.
Hildebrand, S.G. (1971) The effect of coho spawning on the benthic community
of the Platte River, Benzie County, Michigan. Transactions of the American
Fisheries Society, 100, 61–68.
Huggett, A.J. (2005) The concept and utility of ‘ecological thresholds’ in biodiversity conservation. Biological Conservation, 124, 301–310.
Johnston, N.T., MacIsaac, E.A., Tschaplinski, P.J. & Hall, K.J. (2004) Effects of
the abundance of spawning sockeye salmon (Oncorhynchus nerka) on nutrients and algal biomass in forested streams. Canadian Journal of Fisheries and
Aquatic Sciences, 61, 383–403.
Lessard, J.L. & Merritt, R.W. (2006) Influence of marine-derived nutrients
from spawning salmon on aquatic insect communities in southeast Alaskan
streams. Oikos, 113, 334–343.
Lytle, D.A. (2002) Flash floods and aquatic insect life-history evolution: evolution of multiple models. Ecology, 83, 370–385.
Marriott, R.A. (1964) Stream Catalog of the Wood River Lake System, Bristol
Bay, Alaska. United States Fish and Wildlife Service, Washington, DC.
McIntyre, P.B., Jones, L.E., Flecker, A.S. & Vanni, M.J. (2007) Fish extinctions
alter nutrient recycling in tropical freshwaters. Proceedings of the National
Academy of Sciences, 104, 4461–4466.
Merritt, R.W. & Cummins, K.W. (eds) (1996) An Introduction to the Aquatic
Insects of North America, 3rd edn. Kendall. Hall Publishing Co., Dubuque,
IA.
Minakawa, N. & Gara, R.I. (1999) Ecological effects of a chum salmon (Oncorhynchus keta) spawning run in a small stream of the Pacific Northwest. Journal
of Freshwater Ecology, 4, 327–335.
Minakawa, N. & Gara, R.I. (2003) Effects of chum salmon redd excavation on
benthic communities in a stream in the Pacific Northwest. Transactions of
the American Fisheries Society, 132, 598–604.
© 2007 The Authors. Journal compilation © 2007 British Ecological Society, Journal of Animal Ecology, 77, 275–284
284
J. W. Moore & D. E. Schindler
Montgomery, D.R., Buffington, J.M., Peterson, N.P., Schuett-Hames, D. &
Quinn, T.P. (1996) Stream-bed scour, egg burial depths, and the influence of
salmonid spawning on bed surface mobility and embryo survival. Canadian
Journal of Fisheries and Aquatic Sciences, 53, 1061–1070.
Moore, J.W. (2006) Animal ecosystem engineers in streams. Bioscience, 56,
237–246.
Moore, J.W., Schindler, D.E. & Scheuerell, M.D. (2004) Disturbance of freshwater habitats by anadromous salmon in Alaska. Oecologia, 139, 298–308.
Moore, J.W., Schindler, D.E., Carter, J.L., Fox, J., Griffiths, J. & Holtgrieve,
G.W. (2007) Biotic control of stream ecosystem fluxes: spawning salmon
control export of nutrients and matter in Alaskan streams. Ecology, 88,
1278–1291.
Naiman, R.J., Bilby, R.E., Schindler, D.E. & Helfield, J.M. (2002) Pacific
salmon, nutrients, and the dynamics of freshwater and riparian ecosystems.
Ecosystems, 5, 399– 417.
Nilsson, C., Reidy, C.A., Dynesius, M. & Revenga, C. (2005) Fragmentation
and flow regulation of the world’s large river systems. Science, 308, 405–408.
NRC (National Resource Council). (1996) Upstream: Salmon Society in the
Pacific Northwest. National Academy Press, Washington, DC.
Peterson, D.P. & Foote, C.J. (2000) Disturbance of small-stream habitat by
spawning sockeye salmon in Alaska. Transactions of the American Fisheries
Society, 129, 924–934.
Poff, N.L., Allan, J.D., Bain, M.B., Karr, J.R., Prestegaard, K.L., Richter, B.D.,
Sparks, R.E. & Stromberg, J.C. (1997) The natural flow regime: a paradigm
for river conservation and restoration. Bioscience, 47, 769–784.
Polis, G.A., Anderson, W.B. & Holt, R.D. (1997) Toward an integration of
landscape and food web ecology: the dynamics of spatially subsidized food
webs. Annual Reviews of Ecology and Systematics, 28, 289–316.
Power, M.E. (1990) Effects of fish in river food webs. Science, 250, 811–814.
Quinn, T.P. (2005) The Behavior and Ecology of Pacific Salmon and Trout. University of Washington Press, Seattle, WA.
Solan, M., Cardinale, B.J., Downing, A.L., Engelhardt, K.A.M., Ruesink, J.L.
& Srivastava, D.R. (2004) Extinction and ecosystem function in the marine
benthos. Science, 306, 1177–1180.
Steinman, A.D., Lamberti, G.A. & Leavitt, P.R. (2006) Biomass and pigments of benthic algae. Methods in Stream Ecology (eds R. Hauer & G.A.
Lamberti), pp. 357–380. Elsevier Inc., Burlington, MA.
Stockner, J.G., Rydin, E. & Hyenstrand, P. (2000) Cultural oligotrophication:
causes and consequences for fisheries resources. Fisheries (Bethesda), 25, 7–
14.
Taylor, B.W., Flecker, A.S. & Hall, R.O. Jr (2006) Loss of a harvested fish species disrupts carbon flow in a diverse tropical river. Science, 313, 834 –836.
Tiegs, S.D., Chaloner, D.T., Levi, P., Rüegg, J., Tank, J.L. & Lamberti, G.A.
(in press) Timber harvest transforms ecological roles of salmon in southeast
Alaska rain forest streams. Ecological Applications.
Walter, J.K., Bilby, R.E. & Fransen, B.R. (2006) Effects of Pacific salmon
spawning and carcass availability on the caddisfly Ecclisomyia conspersa
(Trichoptera: Limnephilidae). Freshwater Biology, 51, 1211–1218.
Wipfli, M.S., Hudson, J. & Caouette, J. (1998) Influence of salmon carcasses on
stream productivity: response of biofilm and benthic macroinvertebrates in
southeastern Alaska, USA. Canadian Journal of Fisheries and Aquatic Sciences, 55, 1503–1511.
Wright, J.P. & Jones, C.G. (2006) The concept of organisms as ecosystem
engineers ten years on: progress, limitations, and challenges. Bioscience, 56,
203–209.
Received 22 July 2007; accepted 27 September 2007
Handling Editor: Joseph Rasmussen
© 2007 The Authors. Journal compilation © 2007 British Ecological Society, Journal of Animal Ecology, 77, 275–284